TOPSIS using a mixed subjective-objective criteria weights for ABC inventory classification

This paper presents a Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method for ABC inventory classification problems where the inventory are classified - based on multiple criteria - into three groups: A the most important items, B the moderately important ones and C the least important ones. The objective of this classification is to manage and control the inventories in an efficient way based on their weighted scores. To do this, two TOPSIS models using two different distance metrics (first order and second order metrics) and a mixture of subjective-objective criteria weights are proposed. More precisely, this paper addresses the problem of optimizing a set of weights. The objective weights are generated by using the continuous Variable Neighborhood Search (VNS) and the subjective weights are generated by using the Analytic Hierarchy Process (AHP). To test the performance of the proposed models in terms of inventory cost, a benchmark data set commonly used in the relevant literature is exploited.